Design of complex cognitive functions for manipulation using OPRoS sensor components

A robot can be composed of many types of sensors and actuators according to its own use. Moreover, these sensors build up to complex sensor modules in various ways in order to make the robots working in the dynamic environments complete its missions. For the adaptation, it is required to make a robot of which software can be reconfigured easily according to robot missions and working conditions. However, it is very difficult to provide the compatible ability for a manipulator when the robot execute its mission exactly in a structured environment and handle many types of events in dynamic real world. Therefore, we suggest robot software templates which various manipulation functions are achieved based on with slight effort. For this aim, we suggest abstracted sensor model components and state machine components which connect the components with other functional components. Finally, we implement our robot software on multi-core system in order to meet real-time needs including low-latency and strict period time and use a robot middleware, OPRoS (Open software Platform for Robotic Services).